Obstacle-Avoiding Path Planning for Robotic Manipulators Based on Recursive Segmentation Point Migration Optimization and Progressive Inverse Kinematics

Abstract

This paper introduces a novel robotic path planning method that integrates Recursive Segmentation Point Migration Optimization (RSPMO) with Progressive Inverse Kinematics (PIK). RSPMO continuously refines the path by detecting collisions on a direct trajectory, adjusting segmentation points to avoid obstacles, and simplifying the path by removing redundant points, thereby rapidly generating an optimal and efficient trajectory for the manipulator’s end-effector. PIK utilizes second derivative information to iteratively compute joint angles, employing dynamic interpolation and damping factor adjustments near anomalous configurations to enhance system stability. Extensive experimental validation confirms the efficacy of both methods. In a multi-obstacle 2D scenario, RSPMO can plan paths of 789.4516 units in merely 0.0170 seconds, significantly outperforming comparative methods. Additionally, PIK achieves robust performance, with average iteration times of 0.0041 seconds near singular points over 8.45 iterations. Real-world tests on a 6-DOF robotic arm further confirm that the integration of RSPMO and PIK provides an efficient solution for path planning in complex scenarios.